Phase 1 dose-escalation trials
Safety monitoring at this stage is too bespoke and high-touch. Our economics work starting at Phase 2 interventional.
Axiom/AI-NATIVE MONITORING FOR PHASE 2/3 TRIALS
The AI layer for clinical monitoring.
Axiom is an AI-native reasoning layer that reads source data continuously, reconciles against your EDC, and produces sponsor-formatted MVRs your CRAs sign.
$700K of monitoring labor per Phase 2 trial — replaceable by AI
01 / PROBLEM
The gap in modern monitoring - even with eSource.
A typical SMB biotech Phase 2 CRO contract runs $10–15M, with $2.4M of that going to monitoring labor. Roughly 70% of monitoring labor is templated work - source data verification, MVR drafting, query generation, and central oversight.
Your CRAs visit sites every 4–8 weeks and verify data that's already 99.5% correct. The findings that matter - missed AEs, protocol deviations, eligibility errors - wait until the next visit to surface.
That's a structural problem, not a CRA problem. Axiom closes it.
$10–15M
typical SMB biotech Phase 2 CRO contract
Sertkaya et al., 2016
$2.4M
median monitoring line per Phase 2 trial
Industry benchmark · 15–22% of contract
~70%
of CRA monitoring labor is templated work - SDV, MVR drafting, queries, oversight
Axiom analysis · Sertkaya + Tufts CSDD
$700K
per Phase 2 trial replaceable by AI + human-in-the-loop
Andersen 2015 · Abbasi 2025 · ICH E6(R3)
02 / PLATFORM
One reasoning layer. Sits above your existing stack.
Axiom doesn't replace your EDC, your CRO, or your CRAs. It reads source data, reconciles against your EDC, and produces the MVR. Your team signs it.
Reduce monitoring labor 25–40%. Keep quality. Keep your CRO.
Works with your CRO
Runs alongside whichever CRO runs your trial.
Keeps your EDC
Medidata Rave, Veeva Vault, Castor, OpenClinica, IBM CDD.
No site IT burden
FHIR where live, sponsor-mediated exports where not.
4-week deployment
Contract to first MVR.
Per-trial ROI
$2.4M monitoring line · per Phase 2 trial
CRA-signed output
Human-in-the-loop, 21 CFR Part 11 audit trail.
Recovered, per trialEst. range
$400K–$900K
25–40% of Axiom's addressable share of monitoring labor, per trial
Typical CRO contract$10M–$15M
Monitoring line (15–22%)$1.8M–$2.8M
Axiom addressable share (60–75%)$1.1M–$2.0M
Recovery at 25–40% efficiency$400K–$900K
Estimate, Phase 2 interventional. Recovery varies by therapeutic area and trial complexity.
03 / HOW IT WORKS
Three stages, one continuous loop.
Set up once. Run continuously. Ship an MVR your CRA signs.
01 · One-time setup
Protocol & CRF ingestion
Axiom reads your protocol and CRF spec once. The agent extracts every data point, visit window, and prohibited medication into a structured schema - your clin-ops lead verifies it.
Human-verifiedReusable across sites
02 · Continuous
Continuous source data reconciliation
The agent reads new notes, labs, and imaging as they arrive via FHIR where available, and sponsor-mediated exports where not. For each CRF field, it locates source evidence and reconciles against the EDC - flagging discrepancies, missed AEs, and deviations within 48 hours.
FHIR where availableSponsor-mediated exportsCitation-grounded
03 · Weekly output
MVR generation & sign-off
Axiom produces sponsor-formatted Monitoring Visit Reports for CRA review and signature. Every finding cites the source line. TMF-ready, Part 11 audit-trailed, human-in-the-loop.
TMF-ready21 CFR Part 11CRA signs every MVR
04 / DEPLOYMENT
From contract to live monitoring - in four weeks.
Not four months. Not an enterprise IT review. Four weeks of focused work between your clin-ops lead and our deployment engineer.
1
Week 01
Protocol ingested
Schema extracted from your protocol and CRF spec. Verified by your clin-ops lead in a 90-minute review session.
2
Week 02
Data access live
FHIR where available; sponsor-mediated exports where not; EDC read access configured.
3
Week 03
First findings
Agent runs retrospectively on existing enrolled patients. You review the first batch before anything goes into an MVR.
4
Week 04
First MVR delivered
Sponsor-formatted, CRA-signed, filed to TMF. Every finding links to a cited source line.
05 / ECONOMICS
Estimate your recovered spend.
Size your CRO contract and how many Phase 2 trials you run in parallel to translate contract value into implied monitoring line and Axiom’s recovery range.
Per-trial economics.
Model contract size and concurrent trial count to estimate monitoring line, Axiom-addressable share, and expected recovery.
Sources: Sertkaya et al., Br J Clin Pharmacol 2016 (Phase 2 cost data) · Applied Clinical Trials 2024 (monitoring labor share) · ICH E6(R3) (risk-based monitoring guidance). Labor decomposition and efficiency ranges are Axiom's; we show the math below.
Per-trial and portfolio recovery
Monitoring line, per trial$2.3M–$2.9M(18–22% of contract)
Axiom addressable share, per trial$1.6M–$2.0M(70% of monitoring)
Recovery per trial$410K–$800K(25–40% efficiency)
Annual portfolio recovery$1.2M–$2.4M(across 3 concurrent trials)
Directional model. Recovery varies by therapeutic area, trial complexity, and site network. Range across typical Phase 2 trials: $400K–$900K per trial. We'll replace this with a grounded estimate once we align on your portfolio.
About these numbers06 / BUILT FOR
Shaped around the people you already have.
Axiom is a layer - not a replacement for your team, your CRO, or your EDC. It clears routine work so human judgment can land where it matters.
For your CRAs
Augmented, not replaced.
CRAs focus on site relationships, PI engagement, and safety judgment - the work routine SDV displaces. They review and sign every MVR.
For your data
Every finding, cited.
Each flag links to the exact source document line - EHR note, lab result, imaging report. Hallucination rate is measured and published.
For your oversight
Sponsor-owned, audit-ready.
Full audit trail. Full IP retention. 21 CFR Part 11, SOC 2 Type II, HIPAA, GxP documented. You own the data; we run the agent.
07 / FIT
Who this isn't for.
Clarity about what we don't do saves everyone time.
Sponsors with <$250K monitoring budget per trial
At that spend level, the ROI doesn't justify the deployment overhead.
Sponsors running fewer than one Phase 2 trial per year
Axiom pays back over per-trial economics, not per-year licenses.
Teams unwilling to let their CRA sign on AI-drafted output
Axiom is human-in-the-loop by design. If you want fully-automated MVRs, we are not the right fit.
Large pharma running 50+ concurrent global trials
Our current shape is built for biotech-scale operations. Enterprise engagements by referral - reach out if you have a specific opportunity in mind.
08 / EVIDENCE
The data-quality case is settled.
What remains is execution at scale.
“Zero errors that changed any result or conclusion.”
A 2025 retrospective study published in Communications Medicine (Nature portfolio) performed retrospective source data verification across 10,101 eCRFs from the I-SPY COVID platform trial (NCT04488081). The finding: zero errors that changed any conclusion. The trial used centralized monitoring by design.
Abbasi et al., Communications Medicine 5, 444 (2025) · DOI: 10.1038/s43856-025-01126-9
What this means for Axiom: if 100% SDV across 10,000+ fields finds zero consequential errors in a trial using centralized monitoring, the marginal value of sending CRAs on-site to verify each field is near zero. Axiom productizes this insight.
2023
ICH E6(R3) makes risk-based, technology-enabled monitoring the expected standard.
2024
Long-context LLMs process full patient charts in one pass with citation grounding.
Now
Data access is feasible through FHIR where available, with sponsor-mediated exports where FHIR is not live.
Works alongside your stack
- Medidata Rave EDC
- Veeva Vault EDC
- Castor · IBM CDD EDC
- OpenClinica EDC
- Epic · Cerner EHR
- athenahealth EHR
- Okta · Azure AD SSO
Validation
- SOC 2 Type II Audited
- HIPAA BAA available
- 21 CFR Part 11 Documented
- GxP IQ/OQ/PQ
- ICH E6(R3) Aligned
- FHIR R4 Interoperable
Get in Touch
Tell us about your next trial.
Share your study context and monitoring model. We'll follow up on how Axiom fits sponsor clinical operations and clinical monitoring.